the best Linux 2566

Yugo

How can you trust a statistic company that does not know the difference between "market share" and "browsing habits"? There stats only show the percent of people that visits certain sites, NOT market share as they claim. The sample is self selecting and as such should not be considered statistically accurate.

Only people that go to there "exclusive on demand network of small to medium enterprise live stats customers" who ever THEY are. If there "small to medium enterprises" all cater to Microsoft customers and users, then Linux users would be undercounted. Again, the stats should not be considered accurate.

Then, browsers like Konqueror, allow the user to change the information a site sees. If the "exclusive on demand network of small to medium enterprise live stats customers" all have sites that require certain browsers, then other browsers would not be counted. Again, the stats can not be trusted.

Then there is the idea of "unique visitor". How is that determined? By IP address? I would think a Linux user is more likely to have a service that gives them a static IP address, while a windows user more likely to have a dial-up giving a bunch of different IP addresses, again, the Linux user would be undercounted.

Sorry, but there is NOTHING in what you quoted that make me believe that the are talking about "market share" as they claim they are. If you want to make the case that they are talking about something other than market share, then there credibility is out the window. Never mind that their sampling methodology is less than reasonably accurate.

No, sorry, those numbers can not be said to reliably reflect the popularity of Linux on a desktop.

-- Because I am tired of google trolls, I have started blocking all usenet posts from Google. Have fun Ethan, Tina, Maureen, or whatever name you chose to go by.